Drug discovery in the next decade: innovation needed ASAP

Drug Discovery Today Volume 16, Numbers 17/18 September 2011REVIEWSCulture of innovation-ASAP (iASAP): Ask powerful questions;Seek the outliers; Accept defeat; Populate astutely.Drug discovery in the next decade:innovation needed ASAPReviews KEYNOTE REVIEWYoussef L. BennaniVertex Pharmaceuticals, 130 Waverly St., Cambridge, MA 02139, USAPharmaceutical companies must find a better way to increase their outputof truly new drugs for the benefit of patients and for their business survival.Here, I highlight a general perspective from within pharmaceuticalresearch as it pertains to research advances in chemistry, biology,pharmacology, pharmacokinetics and toxicology that, if well integrated,stands to put the industry on a productive path. In addition, I provide acomplementary perspective on the corporate culture aspect of innovation.I also introduce a new concept, termed ‘innovation ASAP’ (iASAP; askingpowerful questions, seeking the outliers, accepting defeat and populatingastutely) and provide support for it using examples of several successfuldrugs.YOUSSEFL. BENNANIYoussef L. Bennaniobtained his undergraduate(BSc) and graduatedegrees (MSc and PhD)in chemistry from theUniversite de Montreal,under the guidance ofProfessor StephenHanessian, andconducted post-doctoral studies at The ScrippsResearch Institute with Professor K. Barry Sharpless.He also holds an MBA from Lake Forest GraduateSchool of Management. In 1993, he joined the pharmaceuticalindustry, as a researcher, working forLigand Pharmaceuticals, Abbott Laboratories andAthersys Inc. He currently serves as Vice President ofDrug Innovation at Vertex Pharmaceuticals, headingboth the discovery chemistry and pharmacokineticsdepartments. He has led programs in immunology,oncology, neurology, metabolic and infectious diseases,resulting in numerous preclinical and clinicalmolecular entities; and has published over 170 papersand patents in the field of drug discovery.IntroductionThe pharmaceutical sector, a cornerstone of the healthcare industry, is undergoing dramaticchange, primarily caused by reduced output of new medicines from research and development(R&D) laboratories, drug pricing pressures, stricter regulatory environments and the overallcurrent economic downturn. This makes demands of all pharmaceutical companies to findbetter ways to increase their output of new drugs, through innovation, to both treat patients andmeet their shareholders’ expectations.This article highlights a general perspective from within pharmaceutical research as it pertainsto research advances in drug discovery [including chemistry, biology, pharmacology, pharmacokinetics(PK) and toxicology] and offers a complementary perspective on the corporate cultureaspects of innovation. A new concept, termed ‘innovation ASAP’ (iASAP: asking powerfulquestions, seeking the outliers, accepting defeat and populating astutely) is introduced andsupported by several successful examples in drug discovery and business in general. The goal ofthis article is to add value to all ongoing efforts aimed at innovation in medicine, as a potentialdriver to revive both the healthcare sector and contribute to the economy in general.The current state of the pharmaceutical industry is undeniably dour. A combination of fallingsuccess rates in the development of innovative therapeutics, pending patent expirations for majordrug classes, in addition to the conditions stated above, have created a perfect storm for theindustry [1–5]. This has resulted in: (i) increasing rates of mergers and acquisitions; (ii) strategicshifts toward generics and emerging markets; (iii) dramatic numbers of job losses over the pastE-mail address: youssef_bennani@vrtx.com.1359-6446/06/$ - see front matter ß 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.drudis.2011.06.004 www.drugdiscoverytoday.com 779

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011Reviews KEYNOTE REVIEWthree years; and (iv) rapid shifts toward the use of lower-costcontract research organizations (CROs) in all aspects of the industry(research, development, manufacturing, among others),coupled with a trend toward decreasing internal R&D budgets.This latter point is of timely importance, because the use ofexternal R&D support tends to perpetuate the same approachesor strategies that got the industry into this state of affairs in the firstplace, and does not involve a real new strategy toward innovativeproducts. The claimed lower costs, through CROs, might not betruly value adding, because outsourced full-time employees comeat a higher cost of management and the potential loss of anopportunity to innovate. As Tom Peters states: ‘You can’t shrinkyour way to greatness’ [6]. Finally, the image of the pharmaceuticalindustry, as a well-respected business that improves health, hasbeen tarnished by price increases as well as news reports of medical,marketing and political lobbying, and direct-to-consumermarketing, all resulting in record fines [7].More pointedly, by chasing ‘low-hanging fruit’, target-basedpharmaceutical research has forced most companies to competeon similar targets and approaches, resulting in duplicative andnonproductive investments. Finally, it appears as though previousinvestments in high throughput screening (HTS), combinatorialchemistry, genomics and proteomics over the past two decades(the very technologies that were supposed to keep the industryfrom the abyss) have yielded a low return on investment. Arguably,these perfect storm conditions also provide numerous opportunitiesand remind the industry that necessity is the mother ofinvention.On the brighter side, over the past two decades, the pharmaceuticalindustry has produced some remarkable medicines to treatseveral conditions, ranging from human immune deficiency virusinfections (HIV) to cardiovascular diseases (CVD), resulting inreduced mortality rates, and rheumatoid arthritis through antitumornecrosis factor (anti-TNF) biologics, to name just a few.Additionally, from the research scientific and operations standpoint,much progress has been achieved in key areas of drugdiscovery, despite the perceived notion of the opposite. Suchadvances, if well integrated, will counter this downturn and setdrug discovery on a more productive and successful course. Here, Iprovide a look forward and offer possible solutions toward a moreintegrative, innovative and, probably more productive R&D environment.So, how can the industry claw its way back from the edge?In chaos lies opportunityA recent survey of the top 25 most innovative companies, encompassingall types of business, did not include a single pharmaceuticalor biotechnology company (http://images.businessweek.com/ss/10/04/0415_most_innovative_companies/index.htm).However, a survey of the top 20 most innovative pharmaceuticalor medical device companies, and endorsed by Wall Street, showeda strong correlation between the value creation of a product, asmeasured by market capitalization, and its novelty, as measured byclear clinical and market superiority (http://www.innovaro.com/pressReleaseFiles/Innovation%20Index%202010%20Q2%205-11-10.pdf). These reported rankings are argued to be a cleardemonstration of the demise of the blockbuster business model, ascurrently practiced by large pharmaceutical companies, given howsparsely these large companies are listed amongst the top 20innovators. From these reports, it is evident that the most innovativecompanies, and thus the most value-creating ones, are thosethat master the art (and science) of understanding necessity ratherthan creating it. Perhaps that it is the multitude of ‘follow-on’s’ or‘me-too’s’ to the blockbuster drugs that are the culprit (vide infra).To set the stage for a discussion on innovation, a definition isnecessary. There are many ways to define innovation but, inessence, it is ‘ideas that ship out’, as such linking fresh thinkingto final products. Others define it as ‘a new match between a needand a solution’, where the novelty can be either in the solution orthe need (all depending on how one defines need); or in a newmarriage of the two [8]. From the innovation–valuation standpoint,one could define the result of innovation as product orservice that both enables a positive return on investment andcreates sustainable valuation for its owners.Arguably, ‘the devil is in the details’ and much high-qualitywork occurs from the inception of an idea all the way through toproduct packaging. The point is that identifying the right necessityor medical need (every company professes to focus on unmetmedical needs, but it all depends on how that is defined) andempowering people to work toward that goal, is the way forward.So how does one navigate through the maze of tribulations, upsand downs all the way to value creation?Me-too drugs are failed innovationsThe (mega) blockbuster business model has had a profound effecton the pharmaceutical industry, both positive (earnings) andnegative (culture). Nobody could argue, in this capitalistic world,against marketing a drug to achieve great revenues. Revenueenables further investment, with the goal of repeating an evengreater achievement. Most major pharmaceutical companies haveadopted this paradigm and invested heavily in R&D over the pasttwo decades, more so than any other industry as a percentage ofsales [9]. This approach has forced R&D-intensive companies tofocus mainly on opportunities with >US$ 1 billion annual marketpotential, which, in return, has created blind spots toward manydiseases and perceived medical needs that did not meet thesereturn levels. In addition, it tended to attract several players intosimilar markets, which led to many years of heavy investments todemonstrate clinical non-inferiority or marginal superiority togain Food and Drug Administration (FDA) approval. The currenttrend to catch the current biologics or bio-similar(s) wave, couldwell backfire in a similar way to small-molecule follow-onapproaches, over the next few decades. Furthermore, commercialfunctions have been in the driver’s seat in clinical trial design,which has led to a business- rather than science-led clinical trialapproach, often leading to failure. Although large revenues weresometimes achieved through this model, it seems to have led theindustry to the edge of a cliff, given how difficult it has been toreproduce it routinely [10]. The argument is that the businessoperates in the ‘outlier space’ and repeats are rare. So, what is anoperationally viable new business model for the next cycle?Revenue versus value creationDrug classes that work through similar general mechanisms, suchas statins (e.g. Lipitor 1 , Zocor 1 , Crestor 1 , Vytorin 1 , Lescol 1 andCaduet 1 ), antiulcerants (e.g. Nexium 1 /Losec 1 , Takepron 1 , Rabeprazole1 and Protonix 1 ) and antipsychotics (e.g. Seroquel 1 ,780 www.drugdiscoverytoday.com

Drug Discovery Today Volume 16, Numbers 17/18 September 2011[(Figure_1)TD$FIG]REVIEWS(a) HOF(b)OOOHHNNOHNHNSOOLipitor®Component of Caduet®NOFHONNHOOHOHN NNOSOCrestor®SOONOFFOFHOOHNNOHOHNLescol®SOONOHNNOOH OSOOHNOZocor®Component of Vytorin®Nexium® Rabeprazole® PANTOPRAZOLE® Takepron/Lansoprazole®OFFReviews KEYNOTE REVIEW(c)FONOHNNOSNNN NNNNNONHNONOSN SNN CIHNHSeroquel® Zyprexa® Risperidone® Zeldox® OAbilify®CICIDrug Discovery TodayFIGURE 1Representative drugs from the statin (a), antiulcerant (b) and antipsychotic (c) classes.Zyprexa 1 , Risperidone 1 , Abilify 1 and Zeldox 1 ; Fig. 1), garneredalmost US$ 25 billion, US$ 20 billion and US$ 20 billion in salesduring 2008, respectively. However, during the period from 2008to the present, the very companies that sell these drugs have seentheir valuations drop by 20–40% (relatively similar trends for mostbig pharmaceutical companies apply from 2001 to 2010). I do notwish or need to highlight any particular company in regards torevenue versus valuation. However, simple research throughhttp://www.wikinvest.com leads to the observation that increasedrevenue or gross margins (over up to 5–10 years) correlate inverselywith corporate value creation as measured by lower stock pricesand, thus, lower market capitalization for most large pharmaceuticalenterprises.In other words, price/earnings ratios for some of the pharmaceuticalcompanies dropped from mid-twenties to high singledigitlevels. For these companies, catching up one another bycompeting in similar markets, rather than investing in innovativeways to treat disease, has clearly back fired. This demonstrates theflaws associated with this business model, which has led to the lossof several tens of thousand jobs. Many associations exist, where afew products (e.g. Epogen 1 , Avastin 1 , Tamiflu 1 , Revlimid 1 , Provigil1 , Depakote 1 and Velcade 1 ; Fig. 2), have driven strong valuecreation for their respective companies, by virtue of their uniquenessand, thus, differentiation.It is now undeniable that increased investments in R&D havenot resulted in increased numbers of new molecular entities(NMEs). Additionally, recent large-scale phase 3 failures (e.g. Axitinib/Pfizer;Elesclomol/Syntha-GSK; AS404/Antisoma-Novartis;Figitumumab/Pfizer; Iniparib/BiPar-Sanofi; Vandetanib/AstraZeneca;and Torcetrapib/Pfizer) and scarcity of late-stage assets[11], coupled with imminent patent expirations for mostblockbuster drugs, bode ill for the future of research-based pharmaceuticalcompanies, and for the industry as a whole.Innovation in drug discovery as a survival necessityThe conundrum associated with innovation is as follows: youknow that you are innovating, when your supervisor says ‘Areyou crazy’? Or, is it: you know you are ‘crazy’ when your supervisorsays ‘Are you innovating’?Innovation is a necessity for improving human prosperity andwell-being. It is imperative that the pharmaceutical industryinvests, engages and manages (with a long-term view and commitmentfrom executive management) its resources to ensure thatnew (not ‘me-too’) and value-creating products emerge from itswww.drugdiscoverytoday.com 781

Drug Discovery Today Volume 16, Numbers 17/18 September 2011REVIEWSaggregate mastery of all the knowledge from the above-mentioneddisciplines, and practical, integrative use of such acumen that willlead the way to truly new medicines. As such, it has a long way to goto address routinely the many pitfalls in moving from a HTS hit to anew chemical entity (NCE), and eventually to marketed NME, in afaster and more reliable manner. Conversely, for future majoradvances, practitioners in all the disciplines mentioned above, willneed to become more knowledgeable in areas currently less familiarto them.Numerous examples can be found within R&D organizationswhere a synthetic chemistry, medicinal chemistry observation or aformulation breakthrough has added value while creating a newdrug (my intent here is to highlight the topic of observation andquestioning, rather than conduct an exhaustive listing of valueaddingobservations). This is particularly applicable when solvingaffinity, selectivity, in vitro or in vivo activity and/or properties,including PK or toxicity issues. Examples of these are retrospectiveanalyses of approved drug properties leading to several rule setsand perspectives, ranging from physicochemical attributes ofdrugs, to formulation sciences and application of continuousimprovement–efficiency models applied from manufacturingand indeed, even to evolving concepts on ‘druggable’ targets[19–30]. So, how does one move forward in medicinal chemistry?Biology-based advancesThe conundrum in biology is the interplay between deductive andinductive biology. Parts of it are well understood and integratedwithin medicinal chemistry and drug discovery, whereas othersremain at bay: Deductive biology: where the whole is broken down intosmaller bits is well managed in current pharmaceutical researchpractice [e.g. gene mapping, protein production, screeningassays, biochemical readouts, soluble protein structure andstructure-based drug design (SBDD)] [31,32]. Inductive biology: or putting the puzzle back together, after aphenotype readout, remains a mixture of art and science and, assuch, is a fertile area for innovation (e.g. membrane andcytosolic protein structural dynamics, signaling nuances,disease biology relevance, compensatory mechanisms and[(Figure_3)TD$FIG] cross-species translation) [33–39].Better tool molecules (e.g. small molecules, peptides, RNA- orprotein-based forms) from chemistry or biology will aid in fasterand better elucidation of newer disease-relevant findings, but onlyin the context of improved biological modeling of human diseasephysiology. As examples, recent advances in G protein-coupledreceptors (GPCR) signaling have begun to offer a glimpse of what ispossible through better appreciation of allosteric modulation[40–42], whereas open–closed states of ion channels [43–47] andprotein-folding dynamics are shaping new understanding of structuralbiology [48]. The emerging field of stem cell biology, throughtherapeutics and small-molecule modulation [49], primary humantissue models [50] (normal and pathological) or three-dimensional(3D)-cultures of the relevant cell contexts, is likely to boost thequest for better medicines. So how is greater precision reached inthe biological sciences?PK-based advancesIt is undeniable that better understanding of the chemical andpharmaceutical properties of newly synthesized molecules leadingto better PK parameters has had a major effect in minimizing theattrition of preclinical and early clinical molecules. Better integrationof solubility, cross-species liver microsomal stability, permeabilityand absorption, as well as cross-species metabolism, havebeen well integrated within the medicinal chemistry community.As examples, the observation that Ritonavir 1 could be used as anexposure booster for Lopinavir 1 (and other drugs) through cytochromeP-450 isoform 3A (CYP3A) inhibition; the observation of ametabolite of a histamine 1 (H1) blocker with biological activityleading to Allegra 1 and Zyrtec 1 (Fig. 3), are just a few examples ofasking the right questions and the power of pharmacologicalobservation. However, prediction of human PK–pharmacodynamic(PKPD) properties has been less than reliable, which pointsto the next area of endeavor for this area of drug discovery [51]. So,how does one raise the PKPD sciences to greater predictabilitylevels in the human setting?Animal pharmacology-based advancesPractitioners in pharmaceutical R&D, and medicinal chemists inparticular, are all familiar with how drugs were invented duringthe 1950s through to the 1990s. The type of question asked thenReviews KEYNOTE REVIEWSOONHHONOONHSNNNHOONHNNHOHONHOHOHOON.HClHOOONNOHClRitonavir® Lopinavir® Allegra® Zyrtec®Drug Discovery TodayFIGURE 3Representative drugs emerging from pharmacological observations.www.drugdiscoverytoday.com 783

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011[(Figure_4)TD$FIG]Reviews KEYNOTE REVIEWFONFZetia®OHOHNOO O NSNNHOViagra®CH 3NNCH 3CH 3Drug Discovery TodayFIGURE 4Representative drugs emerging from newer pharmacological or mechanisticobservations.might have included (as examples): how can we reverse an epilepticshock? (leading to Depakote 1 ; Fig. 2) [52]; how can wereduce cholesterol uptake? (leading to Zetia 1 ) [53]; and the fortuitousfinding that led to a weak antihypertensive becomingViagra 1 (Fig. 4), to name but a few [54].In this area, gene knockout technology has provided supportingevidence for expected phenotypes and increases confidence inparticular target investments [55]. The burgeoning area of biomarkers(previously named phamacogenomics/individual gene mappingor personalized medicine; originally expected more than tenyears ago) promises to further refine disease biology focus, as wellas enabling precise and early clinical readouts and patient stratificationin clinical trials [56].However, one key component of animal model-based thinkingis the admission that we do not know a lot of what we should know(or think we know): currently, several animal disease models arenot predictive of clinical outcome [oncology, immunology, centralnervous system (CNS) and other neurological conditions orpain], whereas other models tend to be more reliable: for example,some viral diseases [e.g. influenza, but neither HIV nor hepatitis-Cviral (HCV) infections]; bacterial and fungal infections; CVDthrough measuring high- and low-density lipoprotein (HDL andLDL, respectively) and simple blood chemistry-based readouts (e.g.glucose or cholesterol). One could argue that, in a typical researchproject, 20–30% of the time is spent fine-tuning molecules to fitthe animal model of disease ‘perfectly’. The cost of which, ifinvested in a ‘fast’ proof-of-concept in the human clinical setting,could easily answer the key question. Pharmaceutical researchaimed at cancer treatment, neurological or psychiatric diseases,in particular, are areas in utmost need of new therapeutics, yet themost underserved from the discovery-to-clinical-success standpoint.As mentioned above, animal models in these disease areasare not predictive, yet regulatory agencies require preclinicalinvestigational new drug (IND) packages to contain in vivo animalefficacy and data based on nonpredictive or nonrelevant diseasemodels. It is sad to see that many companies are giving up on thesediseases rather than tackling the fundamental biological, pharmacologicaland developmental causes of the failures. Perhaps theindustry needs to tackle this very problem to demonstrably affectchange in the success rate of oncolytics, neurological agents orneuroleptics, to name but a few.Given that the goal of animal pharmacology drug discovery is todemonstrate target engagement in living species, with functionalconsequence and minimal adverse effects, further research in morevariants of humanized mice will probably help in this area [57–64].So how, and when, should animal pharmacology be best integrated,if at all, going forward?Toxicology-based advancesCurrently, the greatest contributors to preclinical and clinicalNME failures remain toxicology and translational biology forefficacy. Striking the right therapeutic window, with a safe profileis often a challenge in discovery settings. Much research is beingapplied to bring value from the toxicological standpoint: forexample, screening for mutagens, clastogens and general cellulartoxicities [cytotoxicity panels, hepatocytes, phospholipidosis,transporters and human-ether-a-go-go related gene channel(hERG) blockade], in silico predictions of toxicophores, rapid earlyanimal-based toxicological readouts and multi-species cardiovascularreadouts [65–67]. The field is still plagued by clinical idiosyncratictoxicities, particularly in combination settings where druglevels suffer from output from metabolizing enzymes and theirputative effects on many organs. Much effort is needed to predictaccurately patient safety outcomes through preclinical screening.Therefore, how can the industry, in a similar manner to thatachieved in PK (less of a contributor to clinical drug attrition),maximize translational safety or minimize idiosyncratic toxicityfrom NMEs?So far, I have highlighted the necessity for innovation, theperspective of how investor pressure has led the industry to whereit currently is, and the advances and challenges of all the necessaryscientific disciplines that harbor and feed into medicinal chemistry.Several questions were raised but not answered. The wholeconcept of the culture of innovation in the drug discovery settingremains nebulous, given the multitude of ‘pressure points: investment,competition, science knowledge, time, management’, thediversity of sources of innovation and the still-unpredictablenature of medicinal chemistry, drug discovery and clinical outcomes[68–70].IntegrationDespite the advances, and the challenges ahead in the various drugdiscovery disciplines, their integration into a functional, timelyand productive output is an open question. The above-mentionedstatement by Jenssen captures the essence of a major pitfall ofcurrent drug discovery issues: integration of the above-discusseddisciplines. Some differences occur between academia and industry:for example, in graduate studies, there is an emphasis on theindividual and not the team, and cross-department collaborationsare very new in the academic setting, whereas cross-departmentteamwork is important for success in industry; innovation isgenerally fostered in academia, yet some pharmaceutical organizationstend to redefine the word ‘new’ as ‘me-too’ or ‘me-better’.Additionally, higher education systems are organized aroundscientific disciplines and departments, whereas researchers naturallytend to congregate around similar models in the industrialsetting. Alternatively, the industry tends to be organized arounddiseases or therapeutic areas, with emphasis on centers of excellence.There are few organizational models based around ‘drug784 www.drugdiscoverytoday.com

Drug Discovery Today Volume 16, Numbers 17/18 September 2011REVIEWSdiscovery excellence’ or ‘integrated disease discovery and developmentcenters’ or, better still, ‘chemical biology–pharmacology’,‘pharmaceutics development’, and so on. The point here is thatmuch innovation, speed, productivity are probably lost throughlack of better integrative organizational, cultural and scientificcommunication models. Innovative organizational and managementmodels, along the entire spectrum of R&D, will be key inhelping address current shortcomings [71].The following paragraphs are an attempt to capture some of thecultural attributes that are necessary to maximize the chances fornew products, with particular examples in the small-moleculeR&D setting. I propose iASAP as a set of organizational behaviorsthat could help ignite innovation.Ask powerful questionsThe art of asking powerful questions (why, how and what are morepowerful than when, where or yes/no questions) in any researchsetting leads to inquiry, insight, open ideas and depth of thought. aAsking questions, such as how can one find the best treatment fordiabetes? Why do some people develop schizophrenia or Alzheimer’sdisease? Is more profound than asking, for example, whenare we going to have something that works for diabetes?Asking powerful questions will tend to: (i) stimulate reflectivethinking; (ii) challenge assumptions; (iii) shake dogma; and (iv)generate energy. This simple rule of engagement on what isimportant to work on, why is it important and how to solve it,should help researchers better channel their ideas. As Peter Medawarsaid: ‘any scientist of any age who wants to make importantdiscoveries must study important problems. Dull or piffling problemsyield dull or piffling answers’ [72].Organizations should work hard at, and foster, asking the rightquestions rather than be pulled by marketing demands for me-tooproducts. Although me-too medicines provide a short-term relieffrom the financial pressure of losing market share, they foster aculture of innovation complacency. Organizationally, havingteams of ‘scary-smart’ scientists is necessary, but not sufficientfor innovation: talented scientists need to be channeled intoasking inspiring questions, for the right results to happen. Thisis a fundamental difference between drug hunting (or makingmedicines that work in humans) as opposed to ‘academic work’,where great science might be done, often without immediateapplicability to human pathology.Rules are barriers to innovationCurrently, much effort is being devoted to retrospective databasemining (literature and patents) to encode knowledge and find thecure to pharmaceutical product drought. These exercises rangefrom understanding what constitutes a ‘perfect’ drug space, whatphysicochemical attributes drive good pharmacokinetic outcomes(which does not necessarily translate well across species) throughPKPD predictive modeling and allometric scaling, to minimizingidiosyncratic toxicity and defining the physicochemical propertiesof blood–brain barrier-penetrating compounds [73]. Rules area The word ‘powerful’ can be substituted by: relevant, probing, inquisitive,provocative, incisive, bold, thought-provoking. See Vogt, E.E. et al. The Art ofPowerful Questions: Catalysing Insight, Innovation, and Action; for pdf article:http://www.theworldcafe.com/articles/aopq.pdf.inherently deductive; that is, they break down data, behaviorsor thought processes into discrete measurable bundles of activitythat can then be ordered, thus leading to answers. Although thiscan be value adding, blindly following rules and guidelines canalso lead to ‘blindness’ (whereas scientifically based, these guidelineshave to be taken with a grain of salt). Given that rules comefrom retrospective analyses, they can stifle creativity and serve asbarriers to future breakthrough innovations. Rules, in general,tend to suppress the ‘individual’ and many medicines are ‘individuals’operating in exception or outlier space. Rules- or guidelinesbaseddrug discovery does fit within current understanding ofprocess and, as such, is a manager’s dream. It does not requiremuch creativity to follow a set of guides, or standard operatingprinciples previously worked out to generate data. Is part of theproblem that we, as a society, have learned to live with thecrutches of technology, where everything is in ‘kit’ form andthe underlying principles and processes are no longer understood(or worse) cared about?Another aspect in drug discovery is that of contradictions, if oneonly paid attention to them. As examples: Many anilines are mutagenic yet many successful drugs containanilines, including the best-selling drug worldwide, Lipitor. Excellent team work is at the heart of successful drug discovery,yet success requires the extraordinary contributions of individuals. Productivity is crucial to success, yet quantitative goals aretypically meaningless. Performance is measured on an annual basis and is metricbased, even down to the number of assays run, yet true progressof research is completely independent of either the calendar orsuch metrics.Science should move ahead through both integration of the pastand observation of the unusual, valuing contradictions [74], aswell as directed powerful questioning. So why does the industry,follow sets-of-rules so diligently, rather than explore the moreinnovative outlier space?Seek the outliersOutliers, as defined here, are a composite of key individuals, teams,observations, research approaches, work environment, corporatetolerance for, and reinforcement of, different perspectives, adequatereward systems and the ability to seize those ‘aha’ moments:innovation is all about culture. Outliers are not necessarily individualcontributors; they are observations, unique data relationshipsthat do not fit the hypothesis or dogma, or that cause arethink of assumptions [72].Individual outliers and outlier teamsGood innovators are good problem solvers (who might or mightnot be outliers), skilled in the art of visual observation, dataintegration and listening. Their ability to identify the right need,through powerful questioning, will naturally lead to value-creatingideas, concepts, solutions and, eventually, products. ‘Rational’problem-solving theories leading to practical techniques [75] aresuccessfully used in various industrial settings [76]. However, theconcept of ‘gut feeling’ is also present to some extent in the drugdiscovery environment [77]. It is generally associated with keyindividuals who, for some reason, end up either making the bestReviews KEYNOTE REVIEWwww.drugdiscoverytoday.com 785

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011[(Figure_5)TD$FIG]Reviews KEYNOTE REVIEWOOOH 2NAliskiren®OH OFIGURE 5Drugs with unusual physicochemical characteristics and high medical value.HNNH 2 NNONOOONHOHOOHONNNOOO NHNNOOCylosporin®Drug Discovery Todaymolecules or leading others in the right direction of key problemsolving. Although much effort has been directed toward fittingbinary codes to many pieces of the discovery puzzle throughdatabase mining or knowledge encoding to identify why someindividuals ‘just get it’, the answer remains elusive [78,79].BenchmarkingOne of the key issues currently facing scientists and drug hunters(not all outliers) is the simple fact that they do not know what theydo not know. Too often, the problem is over-intellectualized, andonly the very powerful questions are asked, rather than the ‘wonderquestions’, such as ‘why not?’, ‘have we considered?’ and ‘Iwonder’ about how something works, or ought to or could work? Itis important to remember that most, if not all, scientists thrive onsolving great problems and live to tackle big challenges. Althoughit is important to observe the present and study the past, benchmarkingcan be dangerous in leading to only incrementally inventiveproducts. The drug discovery business, just like every otherone, requires intense teamwork. Teams need not comprise onlyoutliers, but should have some onboard and pay attention to theirperspectives. Benchmarking one’s strategies, projects, teams andefforts is a futile exercise, as it generally leads to duplicative andnoninventive work.Drugs falling under ‘outlier space’Given the current success rate from conception to market, whichcould easily be qualified to be 1%, for most drug projects, one canassume that the drug discovery industry operates in the outlierspace, in general, of what is typically acceptable in many otherindustries.In small-molecule drug discovery, ‘high probability chemicaland/or drug space’ can be defined by the properties that correlatewith successful drugs; whereas ‘outlier space’ can be defined asspace falling outside the 90% cluster between any three or four ofthe following parameters: partition coefficient (Log P), molecularweight (MW), hydrogen bond donor (HBD), hydrogen bondacceptor (HBA), polar surface area (PSA), solubility, potency, permeability,absorption, and so on, in a 3D relationship. In otherwords, ‘outlier space’ defines molecules having one or two keyproperties lying in ‘non-conforming’ space. This space can be‘populated’ by unusually effective drugs, despite their ‘non-fitting’properties, or their pharmacokinetic-disposition profiles ormechanism of action, that fall outside the ‘norm’. Several valuablemedicines tend to fall under the outlier space rather than meet allthe ‘rules’. Here are just a few examples: Aliskiren 1 [Fig. 5; MW, 551; PSA, 146, HBA, 23; human %F(percentage oral bioavailability), 3; T 1/2 , 40 hours; bioavailabilityis limited by low permeability, PGP efflux and first-passhepatic extraction; potent drug on renin, high dose effective,exposure variability tolerated] [80]. Cyclosporin 1 (Fig. 5; MW, 1202; PSA, 287; Log P, 14.3; HBD, 6;orally absorbed by passive diffusion, requires microemulsionformulation, approved and valuable for patients) [81]. All natural product-based drugs [Lovenox 1 , Oxycontin 1 ,Taxotere 1 , Tacrolimus 1 , Pulmicort 1 and Nasonex 1 (quinonesteroids) (Fig. 6); erythromycin-based antibiotics, b-lactambasedantibiotics etc.] fall within this outlier space, given thatmolecules of such beauty or complexity are not usuallydesigned [82–84]. Nature seems to strike the right balancebetween potency, selectivity and physicochemical properties,taking full advantage of metabolizing enzymes and transporterswhen needed. She also takes time to make such exquisitelybeautiful and effective molecules and tests and refines themextensively [85]. Some outlier molecules, which would tend to be avoided incurrent design projects; for example, Singulair 1 , Plavix 1(Fig. 7) and Nexium 1 , which all label proteins in vivo [86];Zetia 1 (contains a b-lactam, not involved in the mechanism ofaction) [87] and Velcade 1 , which contains a boronic acid [88],also form another outlier group. Several molecules, currently undergoing clinical trials, suchTelaprevir 1 (HCV protease inhibitor with MW of 680, peptidomimetic,orally bioavailable and safe and efficacious in patientswith HCV) [89,90]; or the effective cancer therapeutic Lupron 1(a peptide, depot formulation, has high efficacy and results inexcellent financial returns) [91]. Many other examples can befound, including peptide antibiotics (e.g. Cubicin 1 [92]) andantifungals (e.g. Cancidas 1 [93]) (Fig. 8).One could also argue that molecules in the ‘outlier space’ willtend to be harder for follow-on programs and lead the way tomarket with minimal competition (one exception being theATPase proton pump inhibitors). More precisely, researchers tend786 www.drugdiscoverytoday.com

Drug Discovery Today Volume 16, Numbers 17/18 September 2011[(Figure_6)TD$FIG]REVIEWSOONaO ONa SO ONaO OO OOOHOHOO NHOO O OH S OO ONaSO O SONaONaOOnHOONOHOO ONHOHO S ONaOOHOHOOHOHOHHOOHHNOOHOOOHOHHN O HO HOHO O OOOOLovenox® Oxycontin® Taxotere®OOOOOHH OOTacrolimus®OPulmicort®OOOOClHHClOHNasonex®OOReviews KEYNOTE REVIEWROONOOOHOHOOOO OOOR = H; Rapamune®R = C(O)C(CH 3)(CH 2OH) 2; Torisel®Drug Discovery TodayFIGURE 6Representative examples of drugs from natural sources.to seek the perfect prototype as early as the hit-to-lead stage, whichboxes them early on into a set of operational constraints thatmight not help in answering the ‘big’ question originally asked.Some of the reasons for this might be both the classic ones, such asselectivity, toxicity, target validation, biology–clinical translationor, alternatively, that researchers are not good observers of outlierspace. Drug discovery researchers usually average toward the[(Figure_7)TD$FIG]middle or safe ground, and like deductive rules because they areeasy to understand, safe, and anyone can follow them.Sources of innovationWhere does one find problems? Not where answers already exist.As such, fast-follower- or ‘me-too’-type programs do not reallyconstitute a new scientific or medical problem to be solved. ThisHOOCISOCINSOHONOOONNNH 2NNOSingulair®Plavix®Hytrin®Drug Discovery TodayFIGURE 7Examples of highly effective, commercially successful drugs that label proteins in vivo.www.drugdiscoverytoday.com 787

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011[(Figure_8)TD$FIG]Reviews KEYNOTE REVIEWNOH 2 NNHONHOHOHHNNOHTelaprevir®OHNONOHOHHNO OH 2 NHNHNONHOHONHONHHHN N OHOOOHOHOHNHOHNNONHNHONHHNONHONHNHONHOHOHNLupron®CONH 2 OOHNNOHCO 2 HOHONHOHNO O NHNOOHNHHNNH 2H 2 NOHONNN ONH H O HHNHO 2 C HO2 C OO O HO NHOHNN N OHHOONH 2 HO OHOCancidas®Cubicin®Drug Discovery TodayFIGURE 8Representative examples of peptide-based or peptido-mimetic clinical agents or successful drugs.careful selection of endeavor promotes research in areas wherethere is less competition, which means more time to search a widerarea.Szent-Gyorgyi [72] advises one to renew old knowledge, andseveral classic stories come to mind here. For example, rapamycinwas toxic to animals, which put a halt to the project during themid-1970s. Almost 20 years later, the immunosuppressant propertiesof this natural product were biologically better understood,paving the way to two approved drugs in immunology (Rapamune1 ) [94] and oncology (Torisel 1 ) [95]; Thalidomide 1 , originallyapproved as an antiemetic agent, was removed from theclinical setting, but later found approved use in oncology [96];Hytrin 1 (Fig. 7), originally developed as antihypertensive agent,later found use as a treatment for benign prostate hyperplasia [97].Each R&D organization has a wealth of historical, corporateresearch knowledge to tap into, and too much of this is likelyto be going untouched and/or unexamined.By contrast, Pasteur’s method [72] is to find contradictionbetween theory, or dogma, and data. A particularly interestingand potentially valuable measure of discoveries is a contradictionbetween what was expected and the newly generated data, whereone or other must be wrong; in such situations, something usefulcan usually be learned. A classic example is the discovery of Zetia 1 ,where the team was targeting acetyl-coenzyme A acetyltransferase(ACAT) inhibitors; however, no correlation was observed betweencompound inhibitory concentration on ACAT and in vivo efficacy inlowering cholesterol absorption [54]. Following the in vivo efficacystructure–activity relationship (SAR), led to this important discoveryand, subsequently, to the identification of its trans-membranecholesterol transporter protein. The contradiction is obvious in thiscase: shallow in vitro SAR, with inverse correlation to in vivo efficacy.Langmuir’s principle [72] was to simply turn the problem on itshead: ‘Sometimes neither serendipity nor planning cooperateswith the desires of the scientists’. When a desired effect is hinderedby various interfering factors, one should deliberately focus on theundesirable factors so as to exaggerate or understand their badeffects. Perhaps something new will be uncovered as a result. As anexample, erythromycin, one of the first successful macrolide antibiotics,had a side effect that resulted in gastrointestinal motility.It turned out that the culprit was not erythromycin itself, but arearrangement derivative (hemiketal furan) produced, under thelow pH of the stomach, which activated motilin receptors, thusleading to gastrointestinal movement [98]. Capitalizing on this‘side effect’ led to research into motilin agonists for the treatmentof both gastroesophageal reflux disease and diabetic gastroparesis,and the discovery of the motilin receptor [99–101].Outliers can be successfulIn his book Outliers: The Story of Success, Malcolm Gladwell arguesthat success is attributed to exceptional people and those whooperate at the extreme outer edge of what is statistically possible[102]. Identifying coworkers with such skills and providing just theright environment for them to operate will undoubtedly lead tovaluable innovation.788 www.drugdiscoverytoday.com

Drug Discovery Today Volume 16, Numbers 17/18 September 2011REVIEWSAnother analogy can be drawn from the business strategymantra: Blue Ocean Strategy. In the latter, the metaphor of redand blue oceans describes the market universe [103].‘Red Oceans are all the industries in existence today—theknown market space. In the red oceans, industry boundariesare defined and accepted, and the competitive rulesof the game are known. Here companies try to outperformtheir rivals to grab a greater share of product or servicedemand. As the market space gets crowded, prospects forprofits and growth are reduced. Products become commodities,and cutthroat competition turns the oceanbloody. Blue oceans, in contrast, denote all the industriesnot in existence today—the unknown market space,untainted by competition. In blue oceans, demand iscreated rather than fought over. There is ample opportunityfor growth that is both profitable and rapid. In blueoceans, competition is irrelevant because the rules of thegame are waiting to be set. Blue Ocean is an analogy todescribe the wider, deeper potential of market space thatis not yet explored.’ (http://en.wikipedia.org/wiki/Blue_Ocean_Strategy).Applying the above metaphor to drug discovery becomes selfexplanatory,and invites innovators to chart new ‘Blue ocean-likedrug discovery’ maps, where competition is lower, but risk andreward are higher, yet more value-creating (Fig. 9). Although theabove perspective might have a ‘pie-in-the-sky’ sentiment to it,practitioners in drug discovery and development, as well as executivemanagement teams, need to think hard about the future ofwhere their investment activities need to be. Duplication, instrategy and scientific endeavor, of what has happened over thepast two decades will be futile, if not incremental. There remainmany seriously underserved diseases [e.g. cancer (all forms), schizophrenia,depression, multiple sclerosis, Parkinson’s, Alzheimer’s,Huntington’s and amyotrophic lateral sclerosis (ALS)diseases; infectious diseases (influenza, gram-negative bacterialinfections), cystic fibrosis, and many more]; that will require[(Figure_9)TD$FIG]Red oceandrug discoveryBlue oceandrug discoveryDrug Discovery TodayFIGURE 9Representation of iASAP ‘property space’ by analogy to the classic Log P–PSAgraph for ‘drug space’. The X-axis indicates risk/reward, whereas the Y-axisindicates innovation.new approaches. Attacking these diseases, with novelty, patienceand vigor, will undeniably create opportunities for corporationsto reap financial benefits. The red ocean has been tried and it istime to discover new ‘territories’. A complementary perspectivecovering discovery management, corporatesizeandwhyfailureto understand scientists’ psyche and motivational drivers hascontributed to drug R&D failure has recently been published[104].So how do we set ourselves up for success?Scientists must become better observers and listeners. As examples:(i) much creative thinking happens at the more junior levels inorganizations (fresh thinking). Promoting such behavior and findingmechanisms to exploit it rapidly, without middle-managementstifling, is likely to pay dividends; (ii) many mistakes,mishaps, or difficulties occur in the execution of any projectand, if properly observed and astutely questioned (why, how,what), could create a new momentum; (iii) a disgruntled coworkermight be suggest ways to change standard operating procedures;(iv) a repeat-failure might be suggesting repositioning for a newsystem; and (v) an untreated or ill-treated disease signals the needfor an entirely new paradigm, as is the case for Rapamycin (Rapamuneand Torisel) (Fig. 6) and Hytrin (Fig. 7).Accept defeat, then go back‘If you can meet with Triumph and DisasterAnd treat those two imposters just the same’: RudyardKipling, from his poem ‘‘If’’.It might appear that inventiveness is a personality trait, rootedin the brain, and that is part of an inquisitive personality. Thedesire to seek the better, more efficient, ‘never-discovered’approach is generally revered and could be attributed to only afew individuals. However, all inventors will easily admit they‘almost’ quit many times on any given project, and thus potentiallymissing important inventions. In the process of coming upwith a ‘new thing or idea’, one will need to accept failure as par forthe course and not get discouraged. In the business of innovation,one could argue that character, the ability to deal with failure, ismore important than intellect. The challenge here is that mostscientists are trained to persevere (and not accept defeat) in theface of negative research outcome, which makes this a harder caseto deal with.The ability to both accept defeat and manage plans movingforward is essential to keeping goals in check. In The Neuroscience ofScrewing Up, Jonah Lehrer eloquently describes how the ability torecognize when to give up, helps one see challenges in a verydifferent way than before [105]. He cites several examples, includingthe story behind the physics leading the Big Bang theory andawarding Penzias and Wilson the 1978 Nobel Prize in physics.In the drug discovery business, researchers are ‘wrong’ approximately90% of the time on preclinical candidates and on 90% ofthe projects (as measured by NCEs into the clinic). Compoundrelatedattrition owing to PK or toxicology; disconnects betweenbiological target and disease; lack of proper clinical efficacy, targetbasedtoxicity or clinical trial design problems, generally contributeto these failure rates. Additional contributions to this minimalReviews KEYNOTE REVIEWwww.drugdiscoverytoday.com 789

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011Reviews KEYNOTE REVIEWsuccess can be strategic, organizational, tactical or cultural in nature.I do not argue for giving up early, but rather for appreciating the netpositive effect of ‘accepting being wrong’, early. Despite all theplanning and assurance, one can be wrong at any stage. Such anacceptance will lead to mind-openness and the ability to see clearerand hopefully minimize error-repeat. As thoroughly described inthe drug discovery business, the cost of failure is high.The key here is ‘directional problem solving’; are the rightquestion(s) being asked (here it is important to have hypothesesand not just open-ended questions) and are the results moving theteam in the direction that they need to go? When that stopshappening, it is imperative to question whether the team hastruly failed or has reached the limits of its current ability to moveforward: for those scientists who have a hard time letting go, eitheruse the concept of ‘back-burnering’, where the project can simmerbut not distract, or park it altogether.Populate astutelyLiterature on sources of innovation reveals that most influentialhigh-value inventions and/or business ideas were produced by atleast two people (not one, as is commonly believed). Examplesabound in the business, scientific and artistic areas, including HP(Hewlett and Packard), Microsoft (Gates and Allen); Google (Paigeand Brin); Apple (Wozniak, Jobs and Wayne); the Beatles (Lennonand McCartney); DNA (Watson, Crick and Franklin) and DNAsequencing (Maxim, Gilbert and Mirzabeckov); asymmetric epoxidation(Sharpless and Katsuki) and LDL metabolism (Brown andGoldstein) [105]. A study of the personal characteristics of thesepairings (and groupings in research environments) points to thecomplementary nature of those involved and to them havingasked powerful questions that led to their success.Of course, many important inventions were attributed to singleindividuals, but this seems to no longer be the norm, given thecomplexities now involved to bring a differentiated product tomarket and the coordinated teamwork necessary for success. Onthe human resources side, a key component of hiring or forming anew team is striking a fair balance between people who disagree withyou and those that execute orders. Organizations should learn fromthese studies and populate their various innovation incubators withthe appropriate complementary personalities to fuel future innovations.Many of the examples cited above, and probably manyinternally kept stories, add credence to this point of view. Additionally,many pharmaceutical organizations are experimentingwith lower-cost drug discovery models (http://www.slate.com/id/2267688/; http://www.slate.com/id/2267342/, [106]), and scientificcrowd-sourcing approaches, however, the future remains uncertainabout how truly innovative or value-adding these will be.Innovation in the pharmaceutical sector is very different fromthat in many others (particularly non-science based businesses),simply because when researchers embark on a truly new project,they generally have no idea what the final product will ‘look like’.Additionally, equating innovations among the various disciplinesof drug discovery is useless, the sum of all is much greater than thetotal. Finally, research-intensive organizations seeking truly innovativeproducts, need to reward publicly and handsomely trueinnovation and innovators (teams and individuals), rather thanrewarding ‘me-too’ or ‘fast-follower’ products.SummaryThe future of the pharmaceutical industry remains bright, despitethis momentary lapse of appreciation from the business and politicalcommunities. Dramatic advances in chemistry, biology, pharmacology,PK, toxicology and medicine have occurred during thepast 15 years (which is approximately the cycle time from discoveryto market). This, in concert with some nurturing of more innovativecultures within the individual companies, will result in new,improved medicines to fight disease. Innovation is not an averagingdown toward a middle ground of best practice, as is the traditional‘consultation’ mentality, but rather an expression of the individualityand creative strengths of each company. Apple is not Microsoft,yet both are creative and successful in different ways; innovation isnot something to be processed, instead it is to be nurtured, wateredand measured in years and not in quarters or months.In summary, I have shared here my perspective on what behavioralattributes could help move non-innovative cultures towardinnovative ones, and to capitalize on their existing talent to helpreshape the pharmaceutical R&D business toward effective prosperity.By linking the four components of ASAP (ask powerfulquestions; seek the outliers; accept defeat and populate astutely),one has a practical, and potentially useful, working formula tostimulate innovation in the research environment and beyond. Asthe saying goes: ‘culture eats strategy for lunch, every time’.AcknowledgementsThe author wishes to acknowledge the editorial comments,support and valued discussions with many colleagues and friends,namely Raymond Winquist, Michael Briggs, Azin Nezami, ScottBiller, Jeremy Green, James Empfield, Pat Walters, StephenHanessian and Derek Lowe; as well as Janet Kibbee for heradministrative support.References b1 Hughes, B. (2010) 2009 FDA drug approvals. Nat. Rev. Drug Discov. 9, 89–922 Dorren Corbett, J. (2010) Drug approvals slipped in 2010. Wall Street J.3 Cavalla, D. and Minhas, R. (2010) Does R&D pay? Drug Discov. Today 15, 2304 Williams, M. (2011) Productivity shortfalls in drug discovery: contributions fromthe preclinical sciences? J. Pharmacol. Exp. Ther. 336, 3–85 Whitesides, G.M. and Deutch, J. (2011) Let’s get practical. Nature 469, 21–22b Author’s note: it is expected that the reader will see attributes of her/himselfin this brief dialog on Innovation: be neither complacent nor kind on yourselfor your organization; this is about challenging yourself to be better, for yourorganization, and the benefit of all.6 Peters, T. (1999) The Circle of Innovation: You Can’t Shrink Your Way to Greatness.Vintage Books7 Ratner, M. (2010) Crossing the line. Nat. Biotechnol. 28, 1232–12358 Terwiesch, C. and Ulrich, K. (2009) Innovation Tournaments: Creating and SelectingExceptional Opportunities. Harvard Business School Press9 DiMassi, J.A. et al. (2003) The price of innovation: new estimates of drugdevelopment costs. J. Health Econ. 22, 151–18510 DiMassi, J.A. and Faden, L.B. (2011) Competitiveness in follow-on R&D: a race orimitation. Nat. Rev. Drug Discov. 10, 23–2711 Hammer, O. (2011) Lessons learned from Sanofi-Aventis’s phase III failure. Seeking-Alpha http://seekingalpha.com/article/250527-lessons-learned-from-sanofiaventis-s-phase-iii-failure790 www.drugdiscoverytoday.com

Drug Discovery Today Volume 16, Numbers 17/18 September 2011REVIEWS12 Galemmo, R.A., Jr et al. (2005) Memorial issue in honor of Dr. Paul A.J. Janssenpreface. J. Med. Chem. 48, 168613 Nicolaou, K.C. et al. (2000) The art and science of total synthesis at the dawn of thetwenty-first century. Angew. Chem. Int. Ed. 39, 44–12214 Driggers, E.M. et al. (2008) The exploration of macrocycles for drug discovery: anunderexploited structural class. Nat. Rev. Drug Discov. 7, 608–62415 Clark, M.A. et al. (2009) Design, synthesis and selection of DNA-encoded smallmoleculelibraries. Nat. Chem. Biol. 5, 647–65416 Noyori, R. (2009) Synthesizing our future. Nat. Chem. 1, 517 Teague, S.J. et al. (1999) The design of leadlike combinatorial libraries. Angew.Chem. Int. Ed. 38, 3743–374818 Festel, G. (2011) Outsourcing chemical synthesis in the drug discovery processdrug. Drug Discov. Today 16, 23719 Lipinski, C.A. (2000) Drug-like properties and the causes of poor solubility andpoor permeability. J. Pharmacol. Toxicol. Methods 44, 235–24920 Leeson, P.D. and Springthorpe, B. (2007) The influence of drug-like concepts ondecision-making in medicinal chemistry. Nat. Rev. Drug Discov. 6, 881–89021 Johnson, T.W. et al. (2009) Using the golden triangle to optimize clearance andoral absorption. Bioorg. Med. Chem. Lett. 19, 5560–556422 Alelyunas, Y.W. et al. (2010) Experimental solubility profiling of marketed CNSdrugs, exploring solubility limit of CNS discovery candidate. Bioorg. Med. Chem.Lett. 20, 7312–731623 Milletti, F. and Vulpetti, A. (2010) Predicting polypharmacology by bindingsite similarity: from kinases to the protein universe. J. Chem. Inf. Model. 50,1418–143124 Lounkine, E. et al. (2010) SARANEA: a freely available program to mine structure–activity and structure–selectivity relationship information in compound data sets.J. Chem. Inf. Model. 50, 68–7825 Agrafiotis, D.K. et al. (2007) SAR Maps: a new visualization technique for medicinalchemists. J. Med. Chem. 50, 5926–593726 Michino, M. et al. (2009) Community-wide assessment of GPCR structuremodeling and ligand docking: GPCR Dock 2008 participants. Nat. Rev. Drug Discov.8, 455–46327 Wassermann, A.M. et al. (2010) Activity landscape representations for SARanalysis. J. Med. Chem. 53, 8209–822328 Gardner, C.R. et al. (2004) Drugs as materials: valuing physical form in drugdiscovery. Nat. Rev. Drug Discov. 3, 926–93429 Connelly, P.R. et al. (2010) The integrated local CMC service provider: toward adeep economy of pharmaceuticals. Pharmaceut. Outsourcing 309130 Inglese, J. and Auld, D.S. (2009) High throughput screening techniques:applications in chemical biology. In Wiley Encyclopedia of Chemical Biology (Begley,T.P., ed.), Wiley-Interscience http://www.sciencedirect.com/science/article/pii/S136759311000046331 Wagner, B.K. and Clemons, P.A. (2009) Connecting synthetic chemistry decisionsto cell and genome biology using small-molecule phenotypic screening. Curr.Opin. Chem. Biol. 13, 539–54832 Terstappen, G.C. et al. (2007) Target deconvolution strategies in drug discovery.Nat. Rev. Drug Discov. 6, 891–90333 Raddatz, R. et al. (2007) Allosteric approaches to the targeting of G-proteincoupledreceptors for novel drug discovery: a critical assessment. Biochem.Pharmacol. 74, 383–39134 Conn, P.J. et al. (2009) Allosteric modulators of GPCRs: a novel approach for thetreatment of CNS disorders. Nat. Rev. Drug Discov. 8, 41–5435 Leach, K. et al. (2007) Allosteric GPCR modulators: taking advantage of permissivereceptor pharmacology. Trends Pharmacol. Sci. 28, 382–38936 May, L.T. et al. (2007) Allosteric modulation of G protein-coupled receptors. Annu.Rev. Pharmacol. Toxicol. 47, 1–5137 Kenakin, T. (2007) Allosteric agonist modulators. J. Recept. Signal Transduct. Res. 27,247–25938 Kenakin, T. (2004) G-protein coupled receptors as allosteric machines. ReceptorsChannels 10, 51–6039 Kenakin, T. (2005) New concepts in drug discovery: collateral efficacy andpermissive antagonism. Nat. Rev. Drug Discov. 4, 919–92740 Horrigan, F.T. and Aldrich, R.W. (2002) Coupling between voltage sensoractivation, Ca2+ binding and channel opening in large conductance (BK)potassium channels. J. Gen. Physiol. 120, 267–30541 Horrigan, F.T. et al. (1999) Allosteric voltage gating of potassium channels. I. Msloionic currents in the absence of Ca2+. J. Gen. Physiol. 114, 277–30442 Monod, J. et al. (1965) On the nature of allosteric transitions: a plausible model. J.Mol. Biol. 12, 88–11843 Klepeis, J.L. et al. (2009) Long-timescale molecular dynamics simulations ofprotein structure and function. Curr. Opin. Struct. Biol. 19, 120–12744 Korzhnev, D.M. and Kay, L.E. (2008) Probing invisible, low-populated states ofprotein molecules by relaxation dispersion NMR spectroscopy: an application ofprotein folding. Acc. Chem. Res. 41, 442–45145 Henzler-Wildman, K. and Kern, D. (2007) Dynamic personalities of proteins.Nature 450, 964–97246 Selkoe, D.J. (2003) Folding proteins in fatal ways. Nature 426, 900–90447 Chiti, F. and Dobson, C.M. (2006) Protein misfolding, functional amyloid, andhuman disease. Annu. Rev. Biochem. 75, 333–36648 Wells, J.A. and McClendon, C.L. (2007) Reaching for high-hanging fruit in drugdiscovery at protein–protein interfaces. Nature 450, 1001–100949 Ebert, A.D. and Svendsen, C.N. (2010) Human stem cells and drug screening:opportunities and challenges. Nat. Rev. Drug Discov. 9, 367–37250 Chudnovsky, Y. et al. (2005) Use of human tissue to assess the oncogenic activity ofmelanoma-associated mutations. Nat. Genet. 37, 745–74951 Rowland, M. et al. (2010) Physiologically-based pharmacokinetics in drugdevelopment and regulatory science. Annu. Rev. Pharmacol. Toxicol. 51,45–7352 Loscher, W., ed. (1999) Valproate, Milestones in Drug Therapy Series, Springer53 Garcia-Calvo, M. et al. (2005) The target of ezetimibe is Niemann-Pick C1-Like 1(NPC1L1). Proc. Natl. Acad. Sci. U. S. A. 102, 8132–813754 Kloner, R.A. et al. (2001) Effects of sildenafil in patients with erectile dysfunctiontaking antihypertensive therapy. Am. J. Hypertens. 14, 70–7355 Zambrowicz, B.P. and Sands, A.T. (2003) Knockouts model the 100 best-sellingdrugs: will they model the next 100? Nat. Rev. Drug Discov. 2, 38–5156 Goodsaid, F.M. et al. (2010) Voluntary exploratory data submission to the FDA andthe EMA: experience and impact. Nat. Rev. Drug Discov. 9, 435–44557 Terzic, A. and Perez-Terzic, C. (2010) Channelopathies: decoding diseasepathogenesis. Sci. Translat. Med. 2 42ps3758 Strom, S.C. et al. (2010) Chimeric mice with humanized liver: tools for the studyof drug metabolism, excretion, and toxicity. Hepatocytes: Methods Protoc. 640,491–50959 Legrand, N. et al. (2009) Humanized mice for modeling human infectious disease:challenges, progress, and outlook. Cell Host Microbe 6, 5–960 Van Duyne, R. et al. (2009) The utilization of humanized mouse models for thestudy of human retroviral infections. Retrovirology 6, 76–9461 Zhang, B. et al. (2009) Mouse models with human immunity and their applicationin biomedical research. J. Cell. Mol. Med. 13, 1043–105862 Cheung, C. and Gonzalez, F.J. (2008) Humanized mouse lines and theirapplication for prediction of human drug metabolism and toxicological riskassessment. J. Pharmacol. Exp. Therapeut. 327, 288–29963 Macchiarini, F. et al. (2005) Humanized mice: are we there yet? J. Exp. Med. 202,1307–131164 Hopkins, A.L. (2008) Network pharmacology: the next paradigm in drugdiscovery. Nat. Chem. Biol. 4, 682–69065 Shukla, S.J. et al. (2010) The future of toxicity testing: a focus on in vitro methodsusing a quantitative high-throughput screening platform. Drug Discov. Today 15,997–100766 Prestwich, G.D. (2007) Evaluating drug efficacy and toxicology in threedimensions: using synthetic extracellular matrices in drug discovery. Acc. Chem.Res. 41, 139–14867 Hughes, J.D. et al. (2008) Drug properties associated with in vivo toxicologicaloutcomes. Bioorg. Med. Chem. Lett. 18, 4872–487568 Munos, B. (2009) Lessons from 60 years of pharmaceutical innovation. Nat. Rev.Drug Discov. 8, 959–96869 Paul, S.M. et al. (2010) How to improve R&D productivity: the pharmaceuticalindustry’s grand challenge. Nat. Rev. Drug Discov. 9, 203–21470 Agarwal, P. and Searls, D.B. (2009) Can literature analysis identify innovationdrivers in drug discovery? Nat. Rev. Drug Discov. 8, 865–87871 Hoffmann, T. and Bishop, C. (2010) The future of discovery chemistry: quo vadis?Academic to industrial: the maturation of medicinal chemistry to chemicalbiology. Drug Discov. Today 15, 26072 Root-Bernstein, R.S. (1989) Discovering, Inventing and Solving Problems atthe Frontiers of Scientific Knowledge. Harvard University Press Chapter 8,pp. 410–41273 Perola, E. (2010) An analysis of the binding efficiencies of drugs and their leads insuccessful drug discovery programs. J. Med. Chem. 53, 2986–299774 Osono, E. et al. (2008) Extreme Toyota: Radical Contradictions that Drive Success at theWorld’s Best Manufacturer. John Wiley & Sons75 Hua, Z. et al. (2006) Integration TRIZ with problem-solving tools: a literaturereview from 1995–2006. Int. J. Bus. Innovat. Res. 111–128. http://en.wikipedia.org/wiki/TRIZ76 Hamm, S. (2008) Tech Innovations for Tough Times. Bloomberg BusinessweekReviews KEYNOTE REVIEWwww.drugdiscoverytoday.com 791

REVIEWS Drug Discovery Today Volume 16, Numbers 17/18 September 2011Reviews KEYNOTE REVIEW77 Chadwick, A.T. and Segall, M.D. (2010) Overcoming psychological barriers to gooddiscovery decisions. Drug Discov. Today 15, 56178 Ecemis, M.I. et al. (2008) A drug candidate design environment using evolutionarycomputation. IEEE Trans. Evol. Comput. 12, 591–60379 Bonabeau, E. (2003) Don’t Trust Your Gut. Harvard Business Review80 Jensen, C. et al. (2008) Aliskiren: the first renin inhibitor for clinical treatment. Nat.Rev. Drug Discov. 7, 399–41081 Giacomini, K.M. et al. (2010) Membrane transporters in drug development. Nat.Rev. Drug Discov. 9, 215–23682 Chin, Y.W. et al. (2006) Drug discovery from natural sources. AAPS J. 8,E23983 Haefner, B. (2003) Drugs from the deep: marine natural products as drugcandidates. Drug Discov. Today 8, 536–54484 Bade, R. et al. (2010) Characteristics of known drug space; natural products, theirderivatives and synthetic drugs. Eur. J. Med. Chem. 45, 5646–565285 Koehn, F.E. and Carter, G.T. (2005) The evolving role of natural products in drugdiscovery. Nat. Rev. Drug Discov. 4, 206–22086 Potashman, M.H. and Duggan, M.E. (2009) Covalent modifiers: an orthogonalapproach in drug design. J. Med. Chem. 52, 1231–124687 Clader, J.W. (2004) The discovery of ezetimibe: a view from outside the receptor. J.Med. Chem. 47, 1–988 Adams, J. et al. (1998) Potent and selective inhibitors of the proteasome: dipeptidylboronic acids. Bioorg. Med. Chem. Lett. 8, 333–33889 Perni, R.B. et al. (2006) Preclinical profile of Telaprevir, a potent, selective, andorally bioavailable inhibitor of hepatitis C virus NS3-4A serine protease.Antimicrob. Agents Chemother. 50, 899–90990 Maltais, F. et al. (2009) In vitro and in vivo isotope effects with hepatitis C proteaseinhibitors: enhanced plasma exposure of deuterated Telaprevir versus Telaprevir inrats. J. Med. Chem. 52, 7993–800191 Badaru, A. et al. (2006) Sequential comparisons of one-month and three-monthdepot leuprolide regimens in central precocious puberty. J. Clin. Endocrinol. Metab.91, 1862–186792 Steenbergen, J.N. et al. (2005) Daptomycin: a lipopeptide antibiotic for the treatmentof serious gram-positive infections. J. Antimicrob. Chemother. 55, 283–28893 Walsh, T.J. et al. (2004) Caspofungin versus liposomal amphotericin B forempirical antifungal therapy in patients with persistent fever and neutropenia. N.Engl. J. Med. 351, 1391–140294 Vézina, C. et al. (1975) Rapamycin (AY-22,989), a new antifungal antibiotic. J.Antibiot. 28, 721–72695 Rubio-Viqueira, B. and Hidalgo, M. (2006) Targeting mTOR for cancer treatment.Curr. Opin. Investig. Drugs 7, 501–51296 Singhal, S. et al. (1999) Antitumor activity of thalidomide in refractory multiplemyeloma. N. Engl. J. Med. 341, 1565–157197 Lepor, H. et al. (1991) The efficacy and safety of terazosin in the treatment of BPH.Prostate 18, 345–35598 Itoh, Z. et al. (1984) Gastrointestinal motor-stimulating activity of macrolideantibiotics and analysis of their side effects on the canine gut. Antimicrob. AgentsChemother. 26, 863–86999 Brown, J.C. et al. (1973) Motilin, a gastric motor activity stimulating polypeptide:the complete amino acid sequence. Can. J. Biochem. 51, 533–537100 Feighner, S.D. et al. (1999) Receptor for motilin identified in the humangastrointestinal system. Science 284, 2184–2188101 Feighner, S.D. et al. (1999) Cloning and identification of the human motilinreceptor isoforms. PCT Int. Appl. WO 9964436 A1 19991216102 Gladwell, M. (2008) Outliers. Little, Brown & Co.103 Kim, W.C. and Mauborgne, R. (2005) Blue Ocean Strategy: How to Create UncontestedMarket Space and Make Competition Irrelevant. Harvard Business Press104 Knutsen, L.J.S. (2011) Drug discovery management, small is still beautiful: why anumber of companies get it wrong. Drug Discov. Today 10.1016/J.DRUDIS.2011.04.002105 Lehrer, J. (2010) Accept defeat: the neuroscience of screwing up. Wired Mag. http://www.wired.com/magazine/2009/12/fail_accept_defeat/106 Clark, D.E. and Newton, C.G. (2004) Outsourcing lead optimisation – the quietrevolution. Drug Discov. Today 9, 492–500792 www.drugdiscoverytoday.com